Kris ... you hit it ...

> As far as I understand VMWare, it is not an emulator, actually,
> but a type of hypervisor that uses many dirty tricks to provide a
> virtual PC on top of a host PC.  That's why it runs close to the
> host processor's speed for most basic execution.  It's when
> privileges instructions need to be executed that it gets hit
> (the x86 architecture doesn't make virtualization easy).
> So in that sense it has similarities to z/VM.

Right.
Virtualization means that the "guest" runs on the underlying hardware
until some obvious exception kicks it into emulation mode  (such as
a privileged operation, I/O or the like).   Just as virtual memory
is physically held in real storage when operated upon,  so a virtual
machine is executed by the real CPU,  until there is a "fault".

I was looking for serious thoughts on comparisons and concepts,
not trying to start a z/VM -vs- VMware war.   I agree that z/VM is
far beyond VMware in many ways,  and that it gets a huge boost
from the hardare.   I am keenly aware that VMware gives you a GUI.
(Thanks Rich for chiming in.)   I wish that it had a command line.
(I do not know if ESX provides a command line interface.)

This seems especially relevant to Linux because it runs on both
of these HW platforms.   We who know Linux/390,  particularly the
VM part of that group,  are in-tune with the issues and values
presented by a hypervisor.   Now there exists true hypervisor
technology on two very different hardware classes.

The Internet is one of the most significant contributors to Linux.
It is based on common protocols,  where unlike systems communicate
using agreed-upon language.   In the case of hypervisors,  we would
all be well served by agreed-upon language to describe a "machine".
That's what I'm looking for.

Given VMwares "host only" networking,  something like

        hcp def nic eth1

makes conceptual sense.   IT'S JUST AN EXAMPLE,  as were
the other 'hcp' commands in my note which started this thread.
It's an example of what *could* be,  perhaps what *should* be.
The difference between "eth1" syntax and zSeries hex addressing
is one of the issues that should be (ahem) addressed.   The labels
and some terms are different,  but the ideas are the same.

Folks,  this is a golden opportunity made more viable by Linux.
I'm not making a prediction of things yet to come.
I'm suggesting that we make it happen.   Make it happen.

Thanks.

-- R;

Reply via email to